“To understand the nature of the present war—for in spite of the regrouping which occurs every few years, it is always the same war—one must realize in the first place that it is impossible for it to be decisive.”
—George Orwell, 1984
The Intercept—a new digital magazine from First Look Media—recently published its inaugural piece by Glenn Greenwald and Jeremy Scahill detailing the revelation that “the primary method to locate targets for lethal drone strikes” stems from analysis of metadata taken from electronic surveillance and geolocation technologies. While the article raises an important and troubling aspect of the ongoing drone war, it runs the risk of missing the forest for the trees by focusing on the method of collection prior to a drone strike rather than the drone strike itself. Here at the Fourth Estate Watch (FEW), we understand the purpose of the article was to report on drone targeting methods but we also believe this analysis is incomplete. The reason for this is simple: civilians in Yemen, Pakistan, Afghanistan, and other countries where these strikes occur are not necessarily concerned with how the U.S. government obtains information prior to a drone strike—they are concerned with the fact that the strikes are happening in the first place. In short, the problem is not so much the use of metadata for targeting as it is the use of drone strikes in general.
An exhaustive report on the civilian impacts of drones published by the law schools at Stanford & NYU, for example, makes clear that these strikes are often discussed with regard to the drone war’s efficacy. Greenwald & Scahill frame their report in The Intercept in largely the same vein, referring to the metadata collection for targeting as “unreliable” or “fallible,” which may lead an unsuspecting reader to conclude that accuracy and reliability are the only questions here that matter. As the Stanford/NYU report Living Under Drones explains: “[t]hat framing, however, fails to take account of the people on the ground who live with the daily presence of lethal drones in their skies and with the constant threat of drone strikes in their communities” (emphasis ours). And that’s precisely the major—and critical—piece of the drone war that’s missing from The Intercept’s analysis. Whether drones attack via signals intelligence (SIGINT) or human intelligence (HUMINT), the end result for civilians in the vicinity of the strikes remains unchanged: horror and trauma.
Readers unaware of the extreme psychological costs the drone war can exact on civilians may be tempted to focus exclusively on the question of strike efficacy. Western journalists may be interested in the conversation as to whether SIGINT or HUMINT lead to more accurate drone strikes, but we’re confident that Afghan, Yemeni, and Pakistani civilians are not really interested in that conversation. What does the origin of the intelligence matter for those who live in areas where a drone strike can occur at any moment and any time? Jo Becker and Scott Shane of The New York Times have reported that “[d]rones have replaced Guantanamo as the recruiting tool of choice for militants” (emphasis ours). The Living Under Drones report echoes—and clarifies—that theme when it says “[i]mportantly, those interviewed for this report also described how the presence of drones and capacity of the US to strike anywhere at any time led to constant and severe fear, anxiety, and stress, especially when taken together with the inability of those on the ground to ensure their own safety” (emphasis ours). Notice that no one interviewed for the thorough Stanford/NYU report mentioned the origin of drone strike intelligence as a problem—the various civilians interviewed for the report clearly identified the problem as the drones themselves. Indeed, the report cites a Pew Research survey indicating that 97% of Pakistanis view “drone strikes as bad policy” (internal citations omitted).
The reporting in The Intercept may lead an unwary reader to conclude that drone strikes in and of themselves are not necessarily bad policy, but rather that the drone targeting practices should be refined and/or restrained. While that’s a laudable goal and we agree that The Intercept raises crucial questions about the drone program, we think the article is incomplete without mentioning the larger problem with the drone war itself. Greenwald and Scahill critique the metadata targeting with phrases such as “rather than confirming a target’s identity with operatives or informants on the ground” and “the government’s targeting tactics are fundamentally flawed,” but these phrases may lead some readers to walk away believing the only important question is how the U.S. government obtains information relating to drone targeting. Some readers may even conclude that HUMINT is a viable solution here. But the article itself undercuts that argument. In the context of metadata collection and the risk of killing innocents, Greenwald & Scahill quote their anonymous JSOC source as saying “We don’t have people on the ground … as we do where we have a strong foothold, like we do in Afghanistan” (emphasis ours). Elsewhere in the article, however, we are told that “during a single year in Afghanistan—where the majority of drone strikes have taken place—unmanned vehicles were 10 times more likely than conventional aircraft to cause civilian casualties” (emphasis ours). How can one believe HUMINT would lead to more accurate drone strikes when the article itself informs us the country where the U.S. government has a “strong foothold” and plenty of HUMINT has a large ratio of inaccurate attacks? As Jane Mayer has written in The New Yorker, “[drone] strikes are only as accurate as the intelligence that goes into them. Tips from informants on the ground are subject to error[.]”
Here at the FEW, we keep an eye on the media so that readers don’t walk away from reporting with unwarranted conclusions. And while we admire and applaud Greenwald and Scahill’s previous work on related issues, we fear that some incautious readers may conclude from the article that (a) a major question about drones is the accuracy of the intelligence behind their strikes and (b) HUMINT would lead to more accurate strikes. We believe neither proposition is true. As noted above, the article itself—and common sense—indicate that HUMINT (particularly in countries like Yemen, Afghanistan, and Pakistan) may not necessarily be any more reliable than SIGINT when it comes to drone strike targeting. More fundamentally, however, the article glosses over the psychological trauma the drone war can inflict on civilians, regardless of the method of intelligence-gathering used for a strike. Those interviewed in the Stanford/NYU report spoke of a “wave of terror” overwhelming a community at the sound of a nearby drone. Another said that “drones are always on my mind. It makes it difficult to sleep.” The report itself draws a blunt conclusion: “US drone strikes are extremely unpopular in Pakistan.” Notice the Stanford/NYU conclusion is not “US use of metadata for drone targeting is extremely unpopular in Pakistan” but rather that what is unpopular are U.S. drone strikes. Period.
Improved targeting of drone strikes won’t necessarily remedy that problem—the visceral and palpable hatred of drone strikes by those on the receiving end of them should be clear at this point—as a separate report published by Columbia Law School & the Center for Civilians in Conflict emphasizes our main point: “[u]nlike deaths and property loss, which may affect one or more families, the fear associated with covert drone strikes affects nearly everyone in a community” (emphasis ours). Even if one entirely removed metadata analysis from drone strike intelligence analysis, then, that still would not address the question of the psychological fear and trauma that afflict virtually all civilians in countries where the drone war is active. Indeed, the general reaction of many in the establishment media has so far been: does the report in The Intercept indicate we should move more to HUMINT for drone strikes? This perspective ignores the realities on the ground for Afghan, Yemeni, and Pakistani civilians who must live with the fear of a drone strike occurring at any moment for years on end. The question of whether drone-strike intelligence should come from HUMINT or SIGINT may be critical to some Western journalists, but we are skeptical that the question has much relevance for civilians on the ground. We want to conclude with a quote from the Stanford/NYU report, which captures a bit of what it must be like to live in an area at the receiving end of a drone strike: “Drones hover twenty-four hours a day over communities in northwest Pakistan, striking homes, vehicles, and public spaces without warning. Their presence terrorizes men, women, and children, giving rise to anxiety and psychological trauma among civilian communities. Those living under drones have to face the constant worry that a deadly strike may be fired at any moment, and the knowledge that they are powerless to protect themselves.”